Automated vs. manual coding of neuroimaging reports via natural language processing, using the international classification of diseases, tenth revision.

Objective: Natural language processing (NLP) can generate diagnoses codes from imaging reports. Meanwhile, the International Classification of Diseases (ICD-10) codes are the United States' standard for billing/coding, which enable tracking disease burden and outcomes. This cross-sectional study aimed to test feasibility of an NLP algorithm's performance and comparison to radiologists' and physicians' manual coding. Methods: Three neuroradiologists and one non-radiologist physician reviewers manually coded a randomly-selected pool of 200 craniospinal CT and MRI reports from a pool of >10,000. The NLP algorithm (Radnosis, VEEV, Inc., Minneapolis, MN) subdivided each report's Impression into "phrases", with multiple ICD-10 matches for each phrase. Only viewing the Impression, the physician reviewers selected the single best ICD-10 code for each phrase. Codes selected by the physicians and algorithm were compared for agreement. Results: The algorithm extracted the reports' Impressions into 645 phrases, each having ranked ICD-10 matches. Regarding the reviewers' selected codes, pairwise agreement was unreliable (Krippendorff α = 0.39-0.63). Using unanimous reviewer agreement as "ground truth", the algorithm's sensitivity/specificity/F2 for top 5 codes was 0.88/0.80/0.83, and for the single best code was 0.67/0.82/0.67. The engine tabulated "pertinent negatives" as negative codes for stated findings (e.g. "no intracranial hemorrhage"). The engine's matching was more specific for shorter than full-length ICD-10 codes (p = 0.00582x10-3). Conclusions: Manual coding by physician reviewers has significant variability and is time-consuming, while the NLP algorithm's top 5 diagnosis codes are relatively accurate. This preliminary work demonstrates the feasibility and potential for generating codes with reliability and consistency. Future works may include correlating diagnosis codes with clinical encounter codes to evaluate imaging's impact on, and relevance to care.

Prolonged Disability following Re-Exposure after Complete Recovery from Aerotoxic Syndrome: A Case Report.

Aerotoxic syndrome is a rarely recognized and poorly defined illness resulting from exposure to products of oil combustion in commercial aircraft bleed air systems. The pathophysiology of this syndrome is not well understood. A 42-year-old male pilot developed duty-limiting symptoms consistent with aerotoxic syndrome following exposure to airline cabin fumes and recurrence of symptoms following re-exposure despite apparent full recovery from the initial exposure event. Given his persistent symptoms and concern for immediate debilitation/incapacitation if re-exposed, the pilot was deemed medically unqualified for return to commercial piloting duties and a disability determination was made. Consideration of aerotoxic syndrome in aircrew members with unusual or episodic symptoms of neuro/cognitive dysfunction is essential for further characterizing this occupational illness, guiding medical management, and understanding its impact on aviation workforce health.

Comparing Pharmacist-Led Telehealth Care and Clinic-Based Care for Uncontrolled High Blood Pressure: The Hyperlink 3 Pragmatic Cluster-Randomized Trial.

BACKGROUND: A team approach is one of the most effective ways to lower blood pressure (BP) in uncontrolled hypertension, but different models for organizing team-based care have not been compared directly. METHODS: A pragmatic, cluster-randomized trial compared 2 interventions in adult patients with moderately severe hypertension (BP≥150/95 mm Hg): (1) clinic-based care using best practices and face-to-face visits with physicians and medical assistants; and (2) telehealth care using best practices and adding home BP telemonitoring with home-based care coordinated by a clinical pharmacist or nurse practitioner. The primary outcome was change in systolic BP over 12 months. Secondary outcomes were change in patient-reported outcomes over 6 months. RESULTS: Participants (N=3071 in 21 primary care clinics) were on average 60 years old, 47% male, and 19% Black. Protocol-specified follow-up within 6 weeks was 32% in clinic-based care and 27% in telehealth care. BP decreased significantly during 12 months of follow-up in both groups, from 157/92 to 139/82 mm Hg in clinic-based care patients (adjusted mean difference -18/-10 mm Hg) and 157/91 to 139/81 mm Hg in telehealth care patients (adjusted mean difference -19/-10 mm Hg), with no significant difference in systolic BP change between groups (-0.8 mm Hg [95% CI, -2.84 to 1.32]). Telehealth care patients were significantly more likely than clinic-based care patients to report frequent home BP measurement, rate their BP care highly, and report that BP care visits were convenient. CONCLUSIONS: Telehealth care that includes extended team care is an effective and safe alternative to clinic-based care for improving patient-centered care for hypertension. REGISTRATION: URL: https://www. CLINICALTRIALS: gov; Unique identifier: NCT02996565.

Comparison of explanatory and pragmatic design choices in a cluster-randomized hypertension trial: effects on enrollment, participant characteristics, and adherence.

BACKGROUND: Explanatory trials are designed to assess intervention efficacy under ideal conditions, while pragmatic trials are designed to assess whether research-proven interventions are effective in "real-world" settings without substantial research support. METHODS: We compared two trials (Hyperlink 1 and 3) that tested a pharmacist-led telehealth intervention in adults with uncontrolled hypertension. We applied PRagmatic Explanatory Continuum Indicator Summary-2 (PRECIS-2) scores to describe differences in the way these studies were designed and enrolled study-eligible participants, and the effect of these differences on participant characteristics and adherence to study interventions. RESULTS: PRECIS-2 scores demonstrated that Hyperlink 1 was more explanatory and Hyperlink 3 more pragmatic. Recruitment for Hyperlink 1 was conducted by study staff, and 2.9% of potentially eligible patients enrolled. Enrollees were older, and more likely to be male and White than non-enrollees. Study staff scheduled the initial pharmacist visit and adherence to attending this visit was 98%. Conversely for Hyperlink 3, recruitment was conducted by clinic staff at routine encounters and 81% of eligible patients enrolled. Enrollees were younger, and less likely to be male and White than non-enrollees. Study staff did not assist with scheduling the initial pharmacist visit and adherence to attending this visit was only 27%. Compared to Hyperlink 1, patients in Hyperlink 3 were more likely to be female, and Asian or Black, had lower socioeconomic indicators, and were more likely to have comorbidities. Owing to a lower BP for eligibility in Hyperlink 1 (>140/90 mm Hg) than in Hyperlink 3 (>150/95 mm Hg), mean baseline BP was 148/85 mm Hg in Hyperlink 1 and 158/92 mm Hg in Hyperlink 3. CONCLUSION: The pragmatic design features of Hyperlink 3 substantially increased enrollment of study-eligible patients and of those traditionally under-represented in clinical trials (women, minorities, and patients with less education and lower income), and demonstrated that identification and enrollment of a high proportion of study-eligible subjects could be done by usual primary care clinic staff. However, the trade-off was much lower adherence to the telehealth intervention than in Hyperlink 1, which is likely to reflect uptake under real-word conditions and substantially dilute intervention effect on BP. TRIAL REGISTRATION: The Hyperlink 1 study (NCT00781365) and the Hyperlink 3 study (NCT02996565) are registered at ClinicalTrials.gov.

The Future of Occupational and Environmental Medicine.

ABSTRACT: Occupational and environmental medicine (OEM) is an ACGME-accredited preventive medicine specialty focused on work as a social determinant of health and population health. OEM providers recognize and mitigate workplace and environmental hazards, treat resultant injuries and illnesses, and promote the health, wellness and resiliency of workers and communities. Multidisciplinary residency training in clinical medicine, epidemiology, public and population health, toxicology, exposure and risk assessment, and emergency preparedness equips them with the skill set needed for leadership roles in diverse settings. These include clinical practice, academia, corporate settings, and governmental agencies. Despite robust job opportunities, a shortage of formally trained OEM physicians remains and is expected to worsen given a declining number of training programs. We examine root causes of the system-level issues impacting the supply of OEM physicians and potential solutions.

Using advanced racial and ethnic identity demographics to improve surveillance of work-related conditions in an occupational clinic setting.

BACKGROUND: Although racial and ethnic identities are associated with a multitude of disparate medical outcomes, surveillance of these subpopulations in the occupational clinic setting could benefit enormously from a more detailed and nuanced recognition of racial and ethnic identity. METHODS: The research group designed a brief questionnaire to capture several dimensions of this identity and collected data from patients seen for work-related conditions in four occupational medicine clinics from May 2019 through March 2020. Responses were used to calculate the sensitivity and specificity of extant racial/ethnic identity data within our electronic health records system, and were compared to participants' self-reported industry and occupation, coded according to North American Industry Classification System and Standard Occupational Classification System listings. RESULTS: Our questionnaire permitted collection of data that defined our patients' specific racial/ethnic identity with far greater detail, identified patients with multiple ethnic identities, and elicited their preferred language. Response rate was excellent (94.2%, n = 773). Non-White participants frequently selected a racial/ethnic subcategory (78.1%-92.2%). Using our race/ethnicity data as a referent, the electronic health record (EHR) had a high specificity (>87.1%), widely variable sensitivity (11.8%-82.2%), and poorer response rates (75.1% for race, 82.5% for ethnicity, as compared to 93.8% with our questionnaire). Additional analyses revealed some industries and occupations disproportionately populated by patients of particular racial/ethnic identities. CONCLUSIONS: Our project demonstrates the usefulness of a questionnaire which more effectively identifies racial/ethnic subpopulations in an occupational medicine clinic, permitting far more detailed characterization of their occupations, industries, and diagnoses.

Introducing Mass Communications Strategies to Medical Students: A Novel Short Session for Fourth-Year Students.

PROBLEM: The World Health Organization calls on all with quality medical information to share it with the public and combat health misinformation; however, U.S. medical schools do not currently teach students effective communication with lay audiences about health. Most physicians have inadequate training in mass communication strategies. APPROACH: In August 2018, a novel 90-minute class at the University of Minnesota Medical School introduced fourth-year medical students to basic skills for communicating with lay audiences through mass media. Instructors were physicians with experience interacting with the general public via radio, op-ed articles, social media, print media, television, and community and legislative advocacy. After a 20-minute lecture and sharing of instructors' personal experiences, students completed two 30-minute small-group activities. They identified communications objectives and talking points for a health topic, drafting these as Tweets or an op-ed article outline, then presented talking points in a mock press conference with their peers, practicing skills just learned. Pre- and postsurveys documented students' previous engagement and comfort with future engagement with mass media messaging. OUTCOMES: Over 1 week, 142 students participated in 6 separate classes, and 127 completed both pre- and postsurveys. Before the course, only 6% (7/127) of students had comfortably engaged with social media and 14% (18/127) had engaged with traditional media in their professional roles. After the course, students self-reported an increase in their comfort, perceived ability, and likelihood of using specific communications skills to advocate for their patients (all P < .001). NEXT STEPS: The course will be expanded into a 5-session thread for third- and fourth-year medical students spread over 2 years. This thread will include meeting physicians who engage with lay audiences, identifying best practices for mass health communication, identifying bias and misinformation, "dos and don'ts" of social media, and communication skills for legislative advocacy.

Prevalence and Risk Factors of Obesity among Workers in the Oil and Gas Industry of the Niger Delta Region.

Background: Obesity prevalence among oil and gas industry workers over the past decade globally has been on the increase, adversely affecting workers' well-being and productivity. Therefore, necessitates the increased need for both occupational, and public health policymakers to take action. Data on obesity prevalence rates among workers in the oil and gas industry of the Niger-Delta region needed to drive public and occupational health policies regarding obesity is scarce. Therefore, this study aimed to determine the prevalence of obesity and its risk factors among oil and gas industry workers in the Niger Delta Region of Nigeria. Methodology: A cross-sectional study of 280 oil and gas industry workers in the Niger Delta region of Nigeria aged 29 -to 61 years, from November - to - December 2017 was done. A structured survey data of Body Mass Index (BMI) measurements and socio-demographics was collected. Statistical analysis with Chi-Square and Multinomial Logistic Regression tests was used. Results: Results indicated that the Obesity prevalence rate among the oil and gas industry workers of the Niger Delta region was 49.6% (p=0.002). Obesity was significantly associated with a family history of cardiovascular diseases (OR=2.761, p=0.001, 95%CI=1.491-5.112). Conclusion: The Obesity prevalence of 49.6% (p=0.002) among oil and gas industry workers in the Niger-Delta region, is statistically significantly high. Therefore, concretive preventive measures are needed to reduce obesity prevalence among workers.

Cardiorespiratory Fitness of Firefighters: Initial Results of a Multi-Phased Study.

OBJECTIVE: To address which body composition (BC) measures best correlate with cardiorespiratory fitness (CRF) in firefighters and develop a model for accurate CRF estimation compared with traditional methods. METHODS: Career firefighters had body mass index (BMI) and waist circumference (WC) measured in addition to body fat percentage (FM%) by dual-energy x-ray absorptiometry (DXA). CRF as maximum oxygen uptake (VO2max) was estimated by rowing machine and measured by indirect calorimetry in a treadmill exercise test. RESULTS: Fifty two firefighters participated (92.3% men). Univariate correlations with BMI were best with WC. Univariate correlations with VO2max were best with FM%. Obesity classifications by BC measures agreed weakly at best. Multivariate analysis of several variables yielded an improved VO2max estimate (R2 = 0.70). CONCLUSIONS: Fire departments may benefit from more sophisticated measures of BC and CRF to evaluate firefighter fitness.

Design of a pragmatic cluster-randomized trial comparing telehealth care and best practice clinic-based care for uncontrolled high blood pressure.

BACKGROUND: Uncontrolled hypertension is the largest single contributor to all-cause and cardiovascular mortality in the U.S. POPULATION: Nurse- and pharmacist-led team-based care and telehealth care interventions have been shown to result in large and lasting improvements in blood pressure (BP); however, it is unclear how successfully these can be implemented at scale in real-world settings. It is also uncertain how telehealth interventions impact patient experience compared to traditional clinic-based care. AIMS/OBJECTIVES: To compare the effects of two evidence-based blood pressure care strategies in the primary care setting: (1) best-practice clinic-based care and (2) telehealth care with home BP telemonitoring and management by a clinical pharmacist. To evaluate implementation using mixed-methods supported by the RE-AIM framework and Consolidated Framework for Implementation Research. METHODS: The design is a cluster-randomized comparative effectiveness pragmatic trial in 21 primary care clinics (9 clinic-based care, 12 telehealth care). Adult patients (age 18-85) with hypertension are enrolled via automated electronic health record (EHR) tools during primary care encounters if BP is elevated to ≥150/95 mmHg at two consecutive visits. The primary outcome is change in systolic BP over 12 months as extracted from the EHR. Secondary outcomes are change in key patient-reported outcomes over 6 months as measured by surveys. Qualitative data are collected at various time points to investigate implementation barriers and help explain intervention effects. CONCLUSION: This pragmatic trial aims to inform health systems about the benefits, strengths, and limitations of implementing home BP telemonitoring with pharmacist management for uncontrolled hypertension in real-world primary care settings.

Occupational medicine clinical practice data reveal increased injury rates among Hispanic workers.

BACKGROUND: Minnesota has an ethnically diverse labor force, with the largest number of refugees per capita in the United States. In recent years, Minnesota has been and continues to be a major site for immigrant and refugee resettlement in the United States, with a large population of both immigrant and native born Hmong, Hispanic, and East Africans. This study seeks to evaluate the injury risk among the evolving minority workforce in the Minnesota Twin Cities region. METHODS: A retrospective cohort study identifying work-related injuries following pre-employment examinations was performed using electronic health records from a large multi-clinic occupational medicine practice. Preplacement examinations and subsequent work-related injuries were pulled from the electronic health record using representative ICD-10 codes for surveillance examinations and injuries. This study included patient records collected over a 2-year period from January 1, 2015, through December, 2016. The patients in this cohort worked in a wide-array of occupations including production, assembly, construction, law enforcement, among others. RESULTS: Hispanic minority workers were twice as likely to be injured at work compared with White workers. Hispanics were 2.89 times more likely to develop back injuries compared with non-Hispanic workers, and 1.86 times more likely to develop upper extremity injuries involving the hand, wrist, or elbow. CONCLUSION: Clinical practice data shows that Hispanic workers are at increased risk for work-related injuries in Minnesota. They were especially susceptible to back and upper extremity injuries. Lower injury rates in non-Hispanic minority workers, may be the result of injury underreporting and require further investigation.

Use of Clinical Decision Support to Increase Premedication Regimen Homogeneity.

PURPOSE: Patients with prior allergic reactions to iodinated contrast require premedication. This study aimed to increase the homogeneity of premedication orders in such patients. METHODS: A point-of-care (POC) clinical decision support (CDS) alert accompanied by an order set was implemented in the electronic health record (EHR) to notify providers of a prior allergic reaction upon ordering an examination involving iodinated contrast. Premedication regimens were retrospectively compared 11 months pre- and 11 months post-alert implementation, with the different regimens being classified as follows: (1) "preferred" (per ACR recommendations), (2) "nonpreferred" (corticosteroid administered <24 hours before examination, but not per ACR recommendations), or (3) "no premedication." RESULTS: Over 22 months, 22,023 iodinated contrast examinations were performed, 200 (186 intravascular, 12 gastrointestinal/genitourinary, 1 intraarticular, 1 intrathecal) being in patients with a documented iodinated contrast allergy (106 pre-, 94 post-alert deployment). Prealert, 41 of 106 patients (38.7%) received a preferred regimen, 47 (44.3%) received nonpreferred regimens, and 18 (17.0%) received no premedication. Postalert, 58 of 94 patients (61.7%) received a preferred regimen, 21 (22.3%) nonpreferred regimens, and 15 (16.0%) no premedication. After alert initiation, the patients prescribed a preferred regimen significantly increased (Z-score = 3.25, P = .001), but there was no significant difference in the proportion of patients with no premedication (Z-score = -0.02, P = .85). In 2 of 200 patients (1.0%), an allergic reaction occurred, both after POC-CDS alert implementation with a preferred regimen administered. CONCLUSIONS: The homogeneity of premedication regimens significantly increased after the alert's launch. However, the proportion of patients with no premedication did not significantly change.

The merits of MIIC.

Minnesota's immunization information system excels at population health surveillance.

Measurement Tools for Integrated Worker Health Protection and Promotion: Lessons Learned From the SafeWell Project.

OBJECTIVES: To describe (a) a conceptual approach, (b) measurement tools and data collection processes, (c) characteristics of an integrated feedback report and action plan, and (d) experiences of three companies with an integrated measurement approach to worker safety and health. METHODS: Three companies implemented measurement tools designed to create an integrated view of health protection and promotion based on organizational- and individual-level assessments. Feedback and recommended actions were presented following assessments at baseline and 1-year follow-up. Measurement processes included group dialogue sessions, walk-through, online surveys, and focus groups. RESULTS: The approach and measurement tools generated actionable recommendations and documented changes in the physical (eg, safety hazards) and psychosocial (eg, health and safety culture) work environment between baseline and 1-year follow-up. CONCLUSIONS: The measurement tools studied were feasible, acceptable, and meaningful to companies in the SafeWell study.

DIABETIC KETOACIDOSIS CRITICAL CARE PATHWAY IMPLEMENTATION: INCORPORATION INTO EMR SIGNIFICANTLY DECREASES LENGTH OF STAY.

OBJECTIVE: We discuss the implementation and outcomes of a diabetic ketoacidosis (DKA) critical care pathway (CCP) at a 462-bed teaching hospital. METHODS: A multi-disciplinary team implemented a DKA CCP that was translated into 3 computerized physician order entry (CPOE) order sets corresponding to the phases of DKA care. Historical and postintervention data were obtained via automated queries of the electronic medical record (EMR) and further analyzed by manual chart review. RESULTS: Average length of stay decreased from 104.3 to 72.9 hours (P = .0003) after implementation of a DKA CCP. CONCLUSION: Outcome data supports the use of a DKA CCP at our institution. ABBREVIATIONS: DKA = diabetic ketoacidosis CCP = critical care pathway EMR = electronic medical record CPOE = computerized physician order entry ICD-9 = International Classification of Diseases, ninth revision LoS = length of stay SQL = standard query language.

Improving red blood cell orders, utilization, and management with point-of-care clinical decision support.

BACKGROUND: The computerized order for red blood cell (RBC) transfusion within our electronic health record was redesigned with integrated clinical decision support (CDS) to reinforce our restrictive transfusion policy. These changes encouraged 1-unit (1U) RBC orders, clarified hemoglobin (Hb) transfusion triggers, and discouraged unnecessary orders. This study assessed whether these changes resulted in durable effects on provider practices. STUDY DESIGN AND METHODS: The study compared three 1-year subperiods from August 2011 to August 2014, with each year corresponding to a historical control period, preintervention and postintervention years. This study analyzed ratios of 1U versus 2-unit (2U) orders and the absolute rate of RBC orders, units charged, Hb transfusion triggers, repeat transfusion orders, and selected clinical indications both institution-wide and across several subpopulations. RESULTS: Our institution-wide ratio of 1U versus 2U orders increased from 0.50 to 1.20 (p < 0.0001) in the pre- to postintervention subperiods, respectively. The number of units charged per day decreased from 15.68 to 13.53 (p < 0.001), while rates of initial and repeat orders remained stable. Proportion of clinical indications used and mean Hb triggers demonstrated generally positive results. The changes observed between the pre- and postintervention years were far greater than changes between historical control versus preintervention years, reinforcing attribution of results to computerized physician order entry changes. CONCLUSION: Use of computerized orders and CDS encouraged a restrictive transfusion policy, which was highly successful in changing provider practices. We also succeeded in decreasing mean Hb triggers and overall utilization of RBCs. These findings persisted across many subpopulations.

Electronic medical record-based performance improvement project to document and reduce excessive cardiac troponin testing.

BACKGROUND: We assessed the utilization rationale behind provider ordering of cardiac troponin I (cTnI) testing for the diagnosis of myocardial infarction after implementation of a hospital-wide serial order protocol. METHODS: During 2 months in 2013, any request for additional cTnI testing within 30 days of the initial serial cTnI order prompted an electronic health record best practice alert (BPA), which included clinical decision support that could be bypassed by giving a clinical indication. cTnI orders were not limited (timing, number), and upon BPA, trigger data was collected for clinical indications and actions, patient stay (duration, location), International Classification of Diseases, Revision 9 diagnosis, cTnI orders, and timing of cTnI measurements. RESULTS: The BPA was triggered 1477 times by 423 providers who cared for 702 patients. There were a mean of 3.6 cTnI results per patient, 2.1 BPAs per patient, and 1.2 visits per patient. Providers (42% of whom were residents) acknowledged and overrode the BPA 97% of the time. In response to the BPA, 65% of providers selected a prepared rationale: 64% acute coronary syndrome/ST-elevation myocardial infarction/non-ST-elevation myocardial infarction; 30% demand ischemia; and 6% non-ACS myocardial necrosis. Of the remaining 35% of providers, 71% listed no rationale for their additional cTnI orders. Of patients with a BPA, 93% had non-ACS-related primary International Classification of Diseases, Revision 9 diagnosis, and 58% of the time, patients' cTnI results never increased during their stay. In 53% of cases, BPAs were generated by a request for an additional cTnI series when <2 results were available. CONCLUSIONS: Providers largely ignored the BPA that warned of potential overutilization of cTnI testing independent of diagnosis, including ACS.